System for producing short-chain volatile fatty acid by kitchen co-fermentation of sludge

Abstract

The invention discloses a system for producing short-chain volatile fatty acid through kitchen co-fermentation of sludge, and relates to the technical field of biochemical engineering and environmental engineering. Comprising a time phase construction module, a causal playback identification module, a flow field mapping correction module, a metabolism gating feedback module, a structure reconstruction steady-state module and a phase traction control module, and reconstructing a gas-liquid phase difference reference surface according to the real-time flow field response, and establishing a reference surface to calibrate an anti-phase trigger window. According to the method, gas-liquid reverse-phase accurate identification is realized through a high-precision time base line and flow field tracking, a causal chain is constructed to early warn an inverted nucleation area in advance, a steady-state energy base plane is established in combination with three-dimensional flow field correction and metabolism monitoring feedback and driving structure cooperative adjustment, and finally disturbance suppression and path self-resetting are realized through phase traction regulation. The acid production efficiency and the system stability are improved.

Description

Technical Field

[0001] This invention relates to the fields of biochemical engineering and environmental engineering, specifically to a system for co-fermenting sludge and kitchen waste to produce short-chain volatile fatty acids. Background Technology

[0002] The "Sludge-Food Waste Co-fermentation System for Short-Chain Volatile Fatty Acid Production" refers to an integrated biotransformation system that uses municipal sludge and food waste as a mixed substrate. Through a synergistic anaerobic fermentation process, organic matter (such as proteins, fats, and carbohydrates) is decomposed into short-chain volatile fatty acids (such as acetic acid, propionic acid, butyric acid, and valeric acid) under specific reaction conditions. This system typically includes a substrate pretreatment unit, a mixing and reaction unit, a temperature and pH control unit, a stirring and gas separation unit, and an acid collection and online monitoring unit. Its core principle is to adjust the ratio of sludge to food waste, the carbon-to-nitrogen ratio, the reaction temperature, the stirring intensity, and the residence time to promote the metabolic activity of acidifying bacteria and inhibit methanogenic bacteria competition, thus terminating the degradation of organic matter at the acid-producing stage and efficiently accumulating short-chain volatile fatty acids. This type of system not only achieves the reduction and harmless treatment of high-concentration organic waste but also provides important intermediate products for bioenergy, biodegradable materials, and carbon sources for wastewater treatment, representing a crucial technological approach to promoting a circular economy of "waste resource recovery and energy conversion."

[0003] The existing technology has the following shortcomings: In existing technologies, co-fermentation systems for sludge and food waste commonly employ intermittent aeration or micro-aeration and stirring to enhance material mixing and gas-liquid mass transfer. However, under dynamic aeration disturbance conditions, gas velocity and liquid shear rate can easily reverse, causing bubbles to abnormally carry liquid during their ascent, forming a gas-liquid inverted structure. When this inverted structure forms, bubbles accumulate in the upper region of the reactor, liquid circulation is obstructed, and short-chain volatile fatty acids accumulate excessively in localized areas, leading to a sharp increase in acidity. Meanwhile, the lower region of the reactor experiences substrate scarcity and decreased microbial activity due to insufficient material uplift. This type of gas-liquid distribution reversal causes severe imbalances in the metabolic zones within the system, resulting in excessive acidification in the upper layer and stagnation of the lower layer reaction, ultimately leading to decreased fermentation efficiency, fluctuating acid production rates, and loss of long-term operational stability.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a system for co-fermenting sludge with kitchen waste to produce short-chain volatile fatty acids, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a system for producing short-chain volatile fatty acids through co-fermentation of sludge and kitchen waste, comprising a time-phase construction module, a causal playback identification module, a flow field mapping correction module, a metabolic gating feedback module, a structural reconstruction steady-state module, and a phase traction control module: The time phase construction module acquires a nanosecond-level unified time baseline, gas rise rate sequence, and liquid shear gradient trajectory. It reconstructs the gas-liquid phase difference reference surface based on the real-time flow field response and establishes a reference surface to calibrate the anti-phase trigger window. The causal playback identification module establishes a causal playback chain based on the gas-liquid phase difference reference surface, performs time reversal on the aeration rhythm, extracts the bubble rising path and flow velocity deviation, calculates the bubble rising energy density, determines the initial nucleation region of gas-liquid inversion, and generates a dynamic risk index. The flow field mapping correction module constructs a three-dimensional flow field mapping under dynamic risk index constraints, couples the gas-liquid density difference and liquid viscosity gradient to form a drift vector field, analyzes the distribution of the floating loop and the return channel, extracts the flow disturbance center and generates flow correction anchor points. The metabolic gating feedback module collects the upper layer acidity, lower layer substrate concentration and dissolved gas content of the reactor based on the flow correction anchor point, generates a stratified metabolic fingerprint, forms a metabolic stratification map based on the differences in metabolic characteristics, and establishes a threshold triggering gating interval. The structural reconfiguration steady-state module generates an adaptive structural reconfiguration sequence based on the threshold-triggered gating interval, driving the jet guide vane, variable orifice aeration ring and bottom circulation guide to respond collaboratively under time constraints, restoring the continuity of the gas-liquid reflux link and forming a steady-state feedback base surface; The phase traction control module generates a phase conjugate breathing traction command set under the support of the steady-state feedback base plane. Based on the residual density feedback signal and the gating signal, it performs dynamic regulation of time-frequency inverse diffusion suppression, amplitude limiting write-back, self-suppressed interference field and anti-phase perturbation to realize closed-loop control of the system.

[0007] Preferably, the gas-liquid phase difference reference surface reconstruction process is as follows: A stable frequency signal is output by an atomic clock, which is then distributed via phase-locked loop to a gas phase imaging device, a liquid phase velocity measuring device, and a time synchronization line to form a unified time baseline. The signal is then distributed to multiple fixed locations in the reactor via optical fiber to establish a consistent triggering point. Three-dimensional spatial positioning is achieved using laser illumination and a high-precision scale plate. The bubble rising velocity sequence and liquid shear gradient trajectory are obtained by high-speed imaging and particle image velocimetry, respectively. The collected data are aligned according to a unified time baseline. A phase difference analysis relationship is established between the gas rising velocity and the liquid shear velocity within the same voxel. Voxel regions with zero phase difference and consistent change direction are selected to form a gas-liquid phase difference reference surface and calibrate the anti-phase trigger window. Extract the time points, position coordinates, normal vectors, and curvature parameters that repeatedly appear within the reference plane, and establish a standard time anchor point table as a unified reference for flow field playback and structural control.

[0008] Preferably, the dynamic risk index generation steps are as follows: A time-reversal sequence is established based on the gas-liquid phase difference reference surface and time anchor point. The bubble trajectory and liquid shear velocity of multiple cycles are aligned to form a continuous and replayable voxel-level flow field data chain. Within the key analysis section, the velocity vector of each bubble in the voxel is calculated and compared with the liquid velocity field to obtain the relative velocity. The energy density of the bubble path is calculated based on the bubble diameter, surface tension, liquid density and liquid viscosity. Regions with energy density more than twice the average and bubble trajectory overlap exceeding a preset threshold are selected as high-disturbance candidate regions, and dynamic risk scores are generated by combining the relative position of the reference surface and the amplitude of shear velocity fluctuations. High-disturbance voxel regions with stable risk scores over multiple periods are numbered and located, and the initial nucleation region coordinates, risk index and distance to the reference surface are output to generate a dynamic risk index list, which is used as a constraint condition for flow field reconstruction.

[0009] Preferably, in the process of calculating the energy density of the bubble path, the velocity difference of each bubble in each voxel is combined with the corresponding bubble diameter, liquid density, liquid viscosity and surface tension parameters, and a three-point weighted sliding window is used to smooth the velocity curve to improve the accuracy of energy estimation and suppress error fluctuations caused by high-speed disturbances.

[0010] Preferably, the process for generating the flow correction anchor point is as follows: High-risk voxel regions were screened based on the dynamic risk index, a local drift vector field was constructed, and data on gas-liquid velocity difference, density difference, and viscosity gradient were recorded. By tracing the synthetic drift vector path, constructing a three-dimensional uplift and backflow channel, and identifying abnormal areas of path density and velocity change as flow imbalance zones; In the flow imbalance region, cluster strong perturbation voxels, and combine gas-liquid phase difference oscillation, shear rate abrupt change and path overlap density to screen perturbation centers; Extract the geometric center of the disturbance center and its shell stabilization structure, establish flow correction anchor points, and record their voxel numbers, triaxial velocity statistics, and shear distribution; Multiple monitoring points were set around the flow correction anchor point to collect data on acidity, dissolved gas concentration, and liquid phase temperature, constructing a metabolic profile and forming a metabolic spatial reference atlas based on the flow correction anchor point.

[0011] Preferably, the threshold-triggered gating interval establishment steps are as follows: A five-point observation array with equal spacing was set up vertically around the identified flow correction anchor point. At each point, data on acidity, substrate concentration and dissolved gas content were collected, and metabolic data points in a three-dimensional coordinate system were formed by synchronously unifying the time baseline. The data from each observation point are used to construct acidity surface, substrate surface and gas surface, and superimposed to generate a time frame raster of metabolic map to identify metabolic misalignment regions with obvious spatial jumps. Time alignment analysis was performed on acidity abrupt changes, substrate fluctuations and gas concentration anomalies in metabolic misalignment regions, and common time periods of change were extracted to construct dynamic gating threshold intervals. When any observation array data continuously enters the gating threshold range, the structural adjustment action is triggered and the metabolic response results are recorded, based on the gas-liquid path concentration, energy density changes and abrupt changes in flow field direction, and the control parameters for the next cycle are updated.

[0012] Preferably, the steady-state feedback datum formation process is as follows: Based on the dynamic gating interval triggering information, the starting point of structural intervention is set and a disturbance level model is constructed to generate a sequence of structural adjustment instructions with time constraints. The jet guide vanes deployed in the upper middle part are driven in sequence to adjust the opening angle to form a deflection disturbance channel, disperse the bubble path and reduce the path density concentration, thus weakening the upward offset trend. After the guide vane is activated, the diameter of the aeration ring nozzle and the gas injection cycle are controlled according to the set rhythm to construct a discontinuous bubble chain group to reduce the liquid carrying effect. After the guide vanes and aeration rings complete coordinated adjustment, the bottom circulation guide device is activated to construct a tumbling flow field that combines upward and outward expansion, restore the continuity of the gas-liquid return path, and establish a steady-state feedback base.

[0013] Preferably, the phase traction control module generates a phase conjugate traction command set under the support of a steady-state feedback base plane, and performs the following steps based on residual density feedback and gating signals: inverse diffusion suppression, amplitude limiting write-back, interference suppression, and anti-phase perturbation control: Based on the steady-state feedback datum, the spatial coordinates of the disturbance center, the rate of change of energy density and the frequency jump node are collected. The path disturbance information block is constructed and the phase traction command set is generated. The action position, action duration and synchronization target angle are set. Based on the phase traction command set, the perturbation rhythm of the path segment is controlled and the path amplitude, perturbation frequency and leakage energy in the energy channel are monitored simultaneously. The perturbation residual density distribution map is constructed and the reverse diffusion blocking intervention, amplitude correction write-back and frequency interference suppression operations are performed. After the instruction set runs, it periodically scans the energy path and generates residual models of amplitude residual, path misalignment rate and frequency jump rate. If the set threshold is exceeded, it automatically performs rhythm optimization and updates the interference structure parameters to form a closed-loop control network.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention achieves precise positioning of the anti-phase trigger point and traceability of flow field evolution through high-precision acquisition of nanosecond-level time baselines and liquid phase shear trajectories. By establishing a causal chain between bubble rising paths and energy density, it can identify nucleation regions of gas-liquid inversion in advance, preventing their expansion into systemic turbulence. Through three-dimensional vector flow field and density-viscosity coupling analysis, it constructs multi-channel disturbance correction anchor points, providing high-resolution references for monitoring spatial metabolic structures. Furthermore, through dynamic acquisition of differences in metabolic parameters between upper and lower layers, it achieves automatic identification and gating response of the layered metabolic state inside the reactor. At the structural level, it introduces jet guide vanes, variable-orifice aeration rings, and bottom guides in synergistic linkage to establish a stable feedback datum based on the restoration of energy path continuity. Finally, through a phase conjugate traction control mechanism, it achieves rhythmic suppression, phase locking, and path correction of disturbance signals, constructing a self-regulating mechanism that can adapt to dynamic environmental changes. The overall technical solution possesses complete sensing, analysis, response, and control capabilities, improving the accumulation efficiency of short-chain fatty acids, the stability of reactor operation, and the adaptability to high-load disturbance environments during co-fermentation. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a schematic diagram of a system for co-fermenting sludge with kitchen waste to produce short-chain volatile fatty acids, according to the present invention. Detailed Implementation

[0017] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0018] This invention provides, for example Figure 1 The system shown is a co-fermentation system for producing short-chain volatile fatty acids from sludge and kitchen waste, comprising a time-phase construction module, a causal playback identification module, a flow field mapping correction module, a metabolic gating feedback module, a structural reconstruction steady-state module, and a phase traction control module. The time phase construction module acquires a nanosecond-level unified time baseline, gas rise rate sequence and liquid shear gradient trajectory, reconstructs the gas-liquid phase difference reference surface based on the real-time flow field response, establishes a stable reference surface to calibrate the anti-phase trigger window, and forms the time anchor point for subsequent dynamic playback. To accurately identify and dynamically replay the gas-liquid inversion phenomenon during the co-fermentation of sludge and kitchen waste, it is necessary to establish a reproducible unified time baseline, gas rise rate sequence, and liquid shear gradient trajectory at nanosecond-level temporal resolution and three-dimensional spatial scale. Based on this, a gas-liquid phase difference reference surface should be reconstructed, an anti-phase trigger window calibrated, and a high-precision time anchor point formed to provide a basis for subsequent dynamic flow field tracking and structural control. Specifically, the following steps are included: A rubidium atomic clock is used as the primary reference frequency source, outputting a stable 10 MHz frequency signal. This signal is then proportionally distributed into three synchronous pulse signals via a phase-locked loop distributor. These signals are connected to the gas phase imaging device, the liquid phase velocimetry device, and the time synchronization correction circuit, respectively, forming a common trigger reference for the three acquisition devices. This reference pulse is distributed via fiber optic channels to photoelectric trigger terminals at four fixed locations on the reactor: the center of the top, the middle of the symmetrical sidewalls, and the center of the bottom, ensuring spatial consistency of the acquisition starting point. A dual-pulse nanosecond-level laser sheet light source is installed on the outside of the reactor. An optically adjustable laser provides a planar laser illumination strip with a thickness of less than 1 mm and a wavelength of 532 nm, providing continuous illumination of the entire flow field. A high-precision glass-etched calibration plate is fixed to the front of the transparent window outside the reactor. The plate is printed with a sub-millimeter-level scale dot pattern with known spacing, serving as a scale for the correspondence between two-dimensional pixels and physical lengths. The upper and lower calibration plates are vertically positioned using a high-precision micro-displacement bracket, forming a spatial plane for depth correction. The gas phase section introduces a pulsed microbubble generator upstream of the intake pipe. This generator employs a programmable microdiaphragm valve, releasing single bubbles with diameters between 100 and 200 micrometers each time, synchronizing with the time baseline signal at the moment of release. The liquid phase section introduces uniformly dispersed neutral polystyrene spherical particles via a high-pressure injection pump. These particles have a diameter between 15 and 25 micrometers, a density of 1.05 g / cm³, and a refractive index of 1.59, providing a clear contrast to water in terms of light scattering. In the acquisition system, a time response path is defined for each pixel, and the cross-correlation between the aforementioned trigger pulse and the acquisition signal is used to determine the time delay of each acquisition device relative to the reference time zero. Combining the image distortion correction model and pixel coordinate transformation function, the three-dimensional voxel grid coordinates corresponding to the spatial location are obtained, and a unique voxel index number is assigned to each voxel. The final output includes three data items: first, the absolute time value at each time point in the unified time baseline; second, the time delay compensation matrix in the acquisition path; and third, the spatial coordinates and voxel number of each pixel or voxel, achieving a unified time scale and unified spatial reference for the sampled data.

[0019] Using a high-speed imaging device at a sampling frequency of no less than 5000 frames per second, the motion trajectory of the aforementioned standard bubble within the laser illumination sheet is three-dimensionally imaged, with an imaging resolution of no less than 1024×1024 pixels. A feature recognition algorithm in the image sequence is used to identify the bubble edges and locate the bubble's centroid. Image registration methods are used to unify the bubble positions of adjacent frames to the unified time baseline determined in the first step, and the position coordinates of each frame are mapped to a three-dimensional voxel grid to achieve voxel-level position distribution. For the position difference of the same bubble at two consecutive time points, its velocity value in voxel coordinates is obtained by dividing by the time interval. All bubble velocities are then concatenated to form a sequence of gas rise rate changes over time. In the liquid phase, a three-dimensional particle image velocimetry system records the motion trajectory of neutral tracer particles. Double-exposure imaging is used to obtain the particle displacement vector within a short time interval, and the depth direction error is corrected by combining the thickness limitation of the flow field laser illumination to ensure the accuracy of the three-dimensional velocity vector. The velocity vector is mapped to three-dimensional voxels, and the velocity spatial gradient is calculated based on the velocity change rate of adjacent voxels. A sliding time window is used to calculate the temporal rate of change of velocity within a voxel. The spatial distribution and temporal variation curves of the aforementioned shear gradients were normalized to a unified time axis and a unified spatial index to generate the liquid shear gradient trajectory. To ensure measurement robustness, a physical time window of 2 microseconds was introduced, and velocity data were extracted according to the start and end points of a unified time baseline. Weighted median filtering was used to remove anomalous abrupt changes. Thus, two core motion quantities were obtained at the voxel level, nanosecond level, and with temporal and spatial alignment: one is the trajectory of the bubble's rising velocity in the flow field as a function of time; the other is the temporal trajectory of the shear rate of change of the liquid in each voxel in each direction, laying the data foundation for phase difference analysis.

[0020] Using the reactor's gravity direction as the reference axis, two sequences are extracted within each voxel: the first is the time series of the gas rise rate, and the second is the time series of the liquid velocity along the reference axis direction; both are data output and aligned in the second step. The two sequences are differencing at each time point to determine their direction of change, forming a velocity direction change sequence and an acceleration sign sequence. Within each voxel, the phase relationship between the two sequences is determined, i.e., which sequence exhibits sign reversal first, indicating whether it is ahead or behind, and the phase difference time between the two sequences is calculated accordingly. The phase difference time series of all voxels are recorded, and planar connected regions with zero phase difference and consistent phase reversal directions are selected in 3D space to form a draft of the gas-liquid phase difference reference surface. To improve the stability of this surface, the reference surface is averaged over five consecutive aeration cycles, and isolated patches with an area less than 20 voxels are filtered out using 3D connectivity, retaining only the principal surface. This principal surface is the stable reference surface. The time point of each phase reversal is extracted from this surface, and the time period 100 nanoseconds prior to its occurrence is traced back to detect the existence of a persistent velocity shift. If a shift exists, it is marked as the start point of the phase reversal trigger window; otherwise, it is excluded. This operation is repeated to calibrate the end point, obtaining the start and end times of the phase reversal trigger window. The spatial consistency and temporal distribution of the phase reversal trigger window occurrences across multiple cycles are statistically analyzed, and the voxel region and window position with the highest frequency of occurrence are selected and defined as the standard phase reversal window. Within the standard phase reversal window, the window start and end times, the voxel index, the local normal vector direction of the reference surface, and the local curvature value are recorded and compiled into a standard time anchor table. This time anchor serves as the reference input for all subsequent data playback and structural control based on phase reversal tracing, possessing clear spatiotemporal meaning, positional reproducibility, and cycle consistency.

[0021] The causal replay identification module establishes a causal replay chain based on the gas-liquid phase difference reference surface, performs time reversal and replay of the entire aeration rhythm process, extracts the bubble-carrying liquid upward floating path and flow velocity deviation, calculates the bubble rising energy density, locates the initial nucleation region of gas-liquid inversion and generates a dynamic risk index, providing constraints for three-dimensional flow field reconstruction. To conduct source analysis of the gas-liquid inversion triggering process, it is necessary to construct a time-reversal sequence of the entire process based on the established gas-liquid phase difference reference surface and time anchor points. This allows for the replay of bubble rising behavior and identification of flow field disturbance characteristics, thereby extracting the initial nucleation region and forming dynamic risk indicators. Specifically, the following steps are included: Using the obtained gas-liquid phase difference reference surface as the baseline, a set of reference surface data with high spatial consistency across 10 cycles is selected and uniformly mapped to a three-dimensional voxel coordinate system. For each reference surface's corresponding time anchor point, a full-process observation window of 4000 nanoseconds is set, moving backward 3000 nanoseconds and forward 1000 nanoseconds. This time window covers all stages from the initial appearance of the disturbance, the formation of the phase inversion, to the establishment of the inversion. Each time point is accurate to 10 nanoseconds, forming 400 time steps. Using each time step as a node, the bubble velocity field and liquid shear velocity field calibrated in the previous stage are called up, and the bubble trajectory is gradually traced back in the three-dimensional voxel grid, while simultaneously matching the changes in the liquid velocity vector field, completing the temporal playback record of each voxel. In the data structure, each frame contains a timestamp, voxel number, bubble centroid coordinates, gas relative velocity vector, liquid shear velocity gradient, and the distance between the reference surface normal vector direction and the voxel center. By aligning multiple periods of data at the same spatial location with a reference plane, a complete temporal and spatial match is achieved, forming a stable, temporally continuous, and spatially comparable playback chain. Based on this, two data organization methods are established: single-period causal playback sequences and multi-period superimposed and fused sequences. The relative time difference between each moment and the reference plane is marked for subsequent path identification and energy analysis. Compared to existing methods that only collect breakpoint states for "before-after" difference analysis, this step achieves full-field playback in the continuous time dimension for the first time, allowing gas-liquid interface perturbations to be traced back to the initial perturbation point, forming a time-inverted dataset that strictly conforms to the causal chain.

[0022] In the generated time-reversal sequence, the period from 1500 nanoseconds before to 200 nanoseconds after the phase reversal reference is selected as the key analysis segment. Bubbles entering any voxel in each time step are numbered, and the positional change of their center point between consecutive time steps is recorded. By removing the position of the bubble's centroid at adjacent time points, the actual velocity vector of the bubble in the current voxel is obtained. Liquid velocity field data is then retrieved from the same voxel at the same time point, and the velocity difference between the two is calculated to obtain the relative velocity of the bubble in the liquid. Within each voxel, the velocity difference of each bubble is accumulated, and combined with the bubble diameter (determined through high-resolution image recognition), bubble surface tension (known medium parameters), local liquid density, and liquid viscosity, the energy transfer per unit path length of each bubble is calculated using classical equations of motion. Accumulating the energy values ​​of all voxels along the bubble's path yields the total energy release curve of the complete ascent path. The unit energy density of all voxels corresponding to the bubble path is averaged to generate a three-dimensional energy density distribution map, reflecting the intensity of bubble disturbance at different spatial locations before and after phase reversal. To avoid velocity abrupt errors caused by high-speed disturbances, a three-point weighted sliding window is used for path smoothing. Simultaneously, bubble identification errors are controlled to within pixel level through a 3D image comparison mechanism. The energy density field output in this step provides a physical basis for subsequent determination of nucleation regions. Compared to traditional methods that only use the average bubble rising velocity to roughly estimate energy consumption or energy input, this approach enables fine-grained, individual-level energy transfer analysis, revealing key disturbance patterns hidden behind the overall average value.

[0023] In the obtained 3D energy density distribution map, all continuous voxel clusters were searched, and those with a unit volume energy density value greater than twice the average energy density of the entire field for that period were preliminarily screened as energy anomaly accumulation areas. Further, bubble path trajectory maps were overlaid, and the overlap rate of bubble trajectories within each voxel cluster was calculated. Only regions with a trajectory overlap greater than 70% were retained as high-energy, high-perturbation candidate regions. Subsequently, using the relative distance data between the spatial position of the reference surface and the voxel center, high-energy, high-perturbation candidate regions close to the anti-phase reference surface (distance less than 10 voxel units) and located in the region at the leading edge of the reference surface (i.e., the negative direction) were defined as nucleation region candidates. Within these regions, the time series of liquid shear velocity fluctuation amplitudes in the voxels was further extracted, and the fluctuation intensity index was calculated and used as a perturbation stability factor. Finally, the energy density, trajectory overlap rate, shear fluctuation intensity, and spatial position relative to the reference surface were weighted and synthesized to form the dynamic risk score of the voxel. The voxel scores of all candidate regions were normalized to generate a 3D risk index field. Consistency analysis of the risk index was conducted across multiple periods. Voxel regions that appeared and maintained stable scores over more than three periods were defined as the initial nucleation regions of gas-liquid inversion. The voxel numbers, spatial coordinates, average risk index, volatility index, and relative distance to the reference surface of these regions were archived to form a complete dynamic risk index list. This index will serve as the priority evolution path in subsequent 3D flow field reconstruction and also as a reference basis for adjusting the anti-phase warning threshold.

[0024] The flow field mapping correction module constructs a three-dimensional flow field mapping under the constraints of dynamic risk index, couples the gas-liquid density difference and liquid viscosity gradient to form a drift vector field, analyzes the flow imbalance distribution of the floating loop and the return channel, extracts the fluid structure disturbance center and generates flow correction anchor points, and establishes a metabolic monitoring spatial reference. To further identify key flow anomalies leading to gas-liquid inversion, a refined mapping analysis of the three-dimensional flow field inside the reactor was performed under the constraint of a dynamic risk index. This established a gas-liquid drift vector field, identified the center of the disturbance structure, and constructed a metabolic spatial reference baseline. The implementation process included the following steps: Based on the generated dynamic risk index map, voxel regions with a risk index higher than 1.5 times the overall average are selected. These regions are typically concentrated in the upper part of the reactor, near the gas-liquid phase difference reference plane, and are areas with a high probability of inversion formation. These voxel clusters are divided into multiple non-overlapping three-dimensional sub-blocks, each assigned an independent number and three-dimensional index for easy subsequent data tracking and path backtracking. A unified time baseline and spatial voxel grid data are used to retrieve the corresponding time series data within each risk voxel, extracting instantaneous fields containing the following variables: bubble rise velocity vector, liquid velocity vector, liquid shear velocity gradient, gas-liquid relative velocity vector, voxel coordinates, and voxel adjacency structure. Then, based on the medium parameter library, local gas density, liquid density, local liquid temperature, and viscosity are retrieved and spatially interpolated to map the physical property parameters to all selected voxels, forming a voxel-level distribution. After this step, the initial state of the local spatial domain and physical property field under the constraints of the dynamic risk index is formed, with all variables synchronized with the time baseline, providing the boundary and foundation for constructing the coupled drift vector field.

[0025] Within each selected voxel, the gas velocity vector and liquid velocity vector at the current time point are extracted, and their difference is calculated as the local relative velocity. Then, using the difference between gas and liquid densities and based on the projection factor along the gravitational direction, the relative velocity is density-corrected to obtain the true upward displacement vector. Subsequently, the liquid velocity vector field is three-dimensionally differencing to extract the velocity gradients along each axis within the voxel, and the direction and intensity of the viscosity gradient field are calculated using viscosity data. The density-corrected drift vector is superimposed with the viscosity gradient vector to construct a synthetic drift vector containing five variables (position, velocity difference, density difference, viscosity gradient direction, and viscosity gradient intensity), which is recorded in the current voxel. By repeating this operation frame by frame along the time baseline, the dynamic drift field over the entire time series is obtained. This method not only considers bubble dynamics and fluid resistance but also incorporates local property differences into the path construction process. Compared to existing methods that rely solely on velocity fields to construct flow field models, it possesses higher local reflectivity and more complete physical constraints.

[0026] Based on the constructed synthetic drift vector field, several starting points are selected from the highest density continuous cluster region among the risk voxels as the starting coordinates for path tracing. Using a three-dimensional path construction algorithm, starting from each starting point, the path gradually extends along the synthetic vector direction, entering adjacent voxels, and recording the vector direction change, path length, and velocity amplitude at each step. When the path enters the reactor top or sidewall boundary region, it is marked as the endpoint of the upward path. The return path is inverted using the reverse vector from the reactor bottom to the path starting point, and the velocity change and spatial distribution are recorded similarly. After establishing a gas-liquid convection path map across the entire domain, the total number and average length of upward paths are calculated, and the distribution range and closure degree of the return paths are compared to analyze the spatial symmetry and stability of the gas-liquid convection structure. If a region is found to have concentrated upward paths, sparse return paths, and a significantly low path intersection rate, this region is a potential flow imbalance zone with a high risk of inducing gas-liquid inversion. This analytical method overcomes the limitations of traditional two-dimensional cross-sectional analysis, identifying flow link breakpoints in three-dimensional space, significantly improving the accuracy of problem localization.

[0027] Multidimensional analysis was performed on the voxel data of the identified flow imbalance regions to screen out the local areas with the greatest abrupt change in vector direction. Voxels with a direction change angle exceeding 60 degrees and local velocity changes greater than a set threshold were defined as high-risk disturbances. These voxels were clustered, and the internal gas-liquid phase difference fluctuation range, shear velocity change rate, and upward path penetration number were calculated for each cluster. Regions meeting the following three conditions were selected: first, the gas-liquid phase difference exhibited unstable oscillations over multiple cycles with a time span exceeding 300 nanoseconds; second, the shear velocity change rate exceeded 1.8 times the global average within five frames; and third, the path penetration density reached the percentile of its height layer. The set of voxels meeting all the above conditions was defined as the disturbance center. Furthermore, stable flow structures were screened in the region surrounding the disturbance center, characterized by shear velocity changes less than 0.5 times the average and path direction fluctuations less than 15 degrees, constructing a stable outer shell. Using the geometric center of the disturbance center and the stable outer shell as anchor points, the voxel index, spatial coordinates, standard deviation of three-axis velocity changes, and historical path penetration number were recorded. This anchor point will serve as the core basis for establishing spatial reference points. This anchor point setting method, based on both physical disturbance intensity and structural stability, avoids the uncertainties of existing methods that use empirical points or preset symmetrical midpoints, and possesses high reproducibility and physical relevance.

[0028] Based on the established flow correction anchor points, three voxel intervals were selected as monitoring points along the upstream and downstream directions of each anchor point, forming a seven-point observation array. In each observation array, a miniature pH fiber optic sensor, a dissolved gas optical absorption probe, and a liquid phase temperature sensor were installed to measure the pH, dissolved gas concentration, and temperature changes at that location. The sensor acquisition frequency was set to 2000 times per second, and the data was synchronized in real time to a timestamp under a unified time baseline. A local coordinate system was constructed centered on the anchor points, mapping the observed values ​​of each point to metabolic characteristic data points along the main flow direction. By fitting the values ​​of each point in each array, a local metabolic profile was drawn and mapped to the gas-liquid path direction, achieving precise localization of metabolic behavior within the flow structure. Furthermore, by comparing the profiles between different anchor points, a panoramic view of metabolic distribution was constructed for subsequent spatial attribution of phenomena such as metabolic stratification, concentrated acidification, and substrate insufficiency. This anchor-driven metabolic reference structure overcomes the distortion problem of static stratified sampling under flow field disturbance conditions, enabling dynamic and directional monitoring of metabolic behavior, and is a key foundation for supporting intelligent regulation feedback.

[0029] The metabolic gating feedback module collects data on the acidity of the upper layer, the concentration of the substrate in the lower layer, and the content of dissolved gas in the reactor based on the flow correction anchor point, generates a stratified metabolic fingerprint, forms a metabolic hierarchical map based on the differences in metabolic characteristics, establishes a threshold-triggered gating interval, and constructs a real-time feedback channel for structural reconstruction and metabolic regulation. To accurately identify and dynamically regulate metabolic disorders caused by structural perturbations, a multidimensional spatial metabolic observation network needs to be constructed based on the flow correction anchor points established in the previous step. This network generates a hierarchical metabolic map driven by structural response, and a gated feedback mechanism is then built to achieve closed-loop coupling between structural remodeling and metabolic regulation. Specifically, this includes the following steps: Three points with the strongest disturbance stability among the identified flow correction anchor points were selected as representative centers. Five observation points were arranged vertically around each anchor point along the reactor, located 10 mm and 5 mm above the anchor point, and 5 mm and 10 mm below it, forming an equidistant five-point observation array. Each observation point was equipped with three types of independent sensors: first, a miniature fiber optic acidity sensor (less than 5 cm in length, with a response time of less than 1 second) to record real-time pH changes in the liquid phase at that point; second, a mid-infrared spectral substrate probe (4.5 μm to 9.5 μm band) to monitor carbon source concentration changes, exhibiting substrate molecular-level selectivity and a response time of no more than 2 seconds; and third, an electrochemical dissolved gas probe covering the three main components—methane, hydrogen, and carbon dioxide—with a detection limit below 0.05 mg / L. The sensors used the anchor point coordinates as a reference to perform three-dimensional spatial annotation of the observation point data and synchronized it to an absolute timestamp under a unified time baseline, ensuring a one-to-one correspondence between the data in time and space. All data are input into the data acquisition card in real time via photoelectric conversion and analog-to-digital conversion modules, and then uploaded to the spectrum construction and processing sequence. Unlike the existing method of setting up acidity meters or sampling ports at single points at the top or bottom of the reactor, this method adopts a five-point array design with the anchor point as the center and equal spacing, which not only ensures the representativeness of the metabolic profile, but also ensures the directional consistency and spatial resolution of the response measurement.

[0030] pH data acquired synchronously from five observation points were normalized, using a relative neutral pH of 7 as the baseline, and the degree of offset at each point was marked. Substrate concentration data were converted to mass concentration in milligrams per liter and smoothed using a three-point moving average within each time frame to eliminate local sampling noise. The volume distribution ratios of the three gas components per unit volume were extracted from the dissolved gas concentration data and then converted into relative offset indices with the trend of change above and below the anchor point as the main axis. Using each frame timestamp as an index, numerical distribution surfaces for the five observation points were constructed in a three-dimensional coordinate system, corresponding to the acidity surface, substrate surface, and gas surface, respectively.

[0031] Then, by overlaying the positional change trends of each metabolic surface frame by frame according to the time series, a complete "metabolic atlas time frame raster" is generated, which records the changes of metabolic parameters over time and space. At each moment, a three-dimensional metabolic profile of the region where the anchor point is located can be obtained, including three dimensions: acidity, substrate, and gas indicators. In this atlas, if the pH value of the upper layer is significantly lower than that of the lower layer, and the substrate concentration rises abnormally at the bottom, and the carbon dioxide concentration in the gas component accumulates rapidly, forming an image boundary with a significant distribution fault, then it can be preliminarily determined that there is a metabolic fault in this region. This metabolic fault is characterized by a triple coupling of acidification upward movement, substrate retention, and uneven gas production. It has obvious spatial hierarchical jumps in the atlas and highly coincides with the path distribution of the flow disturbance center, verifying the spatial coupling relationship between the metabolic fault and local disturbance. This data integration method has not been disclosed in existing technologies. Traditional methods mostly set warning values ​​based on average values ​​and individual parameters, which are difficult to reveal the spatial characteristics and coupling trends of metabolic changes.

[0032] In the constructed 3D metabolic map, local slope fitting and maximum offset analysis were performed on each data time series to extract regions of abrupt pH decreases, substrate concentration fluctuations, and critical regions of gas concentration increases. The time points of significant changes in these three indicators on the time axis were aligned, their common occurrence time periods were determined, and their intersection was taken as the "high-intensity metabolic disturbance segment." Within this segment, if the change in any metabolic indicator exceeds ±30% of the moving average of the past ten time frames and lasts for more than 5 seconds, the metabolic misalignment is considered to have developed into a stable abnormal state. Based on this, a dynamic gating threshold interval was constructed, centered on the high-intensity disturbance segment, with a buffer period of 1 second before and after, setting the effective duration of the metabolic disturbance state between 3 and 8 seconds. If the observation point data re-enters this interval, the gating condition is considered met, triggering the subsequent structural response judgment logic. This gating strategy no longer relies on statically set absolute numerical thresholds but instead captures segments and judges trends based on historical change trajectories, greatly improving the system's sensitivity to real disturbances and its resistance to false alarms. This gating strategy differs fundamentally from traditional upper and lower limit judgment methods based on fixed warning values. It is a dynamic structure discrimination method with memory and evolutionary judgment capabilities.

[0033] When three consecutive frames of metabolic observation array data from a certain anchor point enter the pre-defined gated interval, the system immediately retrieves real-time gas-liquid flow path data and current energy density field information to determine the correlation between the gated event and the flow disturbance direction, path concentration, and flow field imbalance intensity. If all three simultaneously show an increasing trend—that is, an increase in gas path concentration, an increase in energy density, and an increase in the rate of change of path direction—a feedback command is sent to the structural adjustment actuator. This command includes the aeration ring orifice angle adjustment value, the target position of the flow guide structure, and the bottom circulation pressurization duration parameter.

[0034] After the action is executed, the metabolic response is immediately monitored. If the monitoring data shows an increase in acidity, recovery of substrate concentration, or a tendency for gas composition to reach equilibrium, the adjustment effect is recorded and the upper and lower limits of the gate threshold for the next cycle are updated. If the monitored values ​​do not show significant improvement, the structural parameters are adjusted a second time, and the control matrix is ​​updated to ensure the intervention remains effective. The entire process, from monitoring to judgment to execution to verification to parameter tuning, forms a complete closed loop, ensuring that structural disturbances respond reversibly to changes in metabolic stratification. Unlike previous reactor control methods that rely solely on time-based timing or fixed control logic using flowcharts, this method constructs a control channel that can dynamically respond to changes in flow structure through a metabolic spatial structure mapping and time-triggered coupling mechanism, exhibiting significant advantages in directionality, adaptability, and adjustment precision.

[0035] The structural reconfiguration steady-state module generates an adaptive structural reconfiguration sequence based on the threshold-triggered gating interval, driving the jet guide vane, variable orifice aeration ring and bottom circulation guide to respond collaboratively under time constraints, weakening the gas-liquid inversion formation mechanism, restoring the continuity of the gas-liquid reflux link, and forming a steady-state feedback base surface after the energy transmission channel stabilizes, providing a reference input for dynamic traction control; To address the structural disturbance imbalance exposed by the metabolic misalignment triggering gating region, a temporal structural regulation sequence needs to be generated based on dynamic triggering information. This sequence then drives the gas-liquid disturbance regulation components within the reactor to restore the continuity of the gas-liquid reflux path, construct a stable energy channel, and ultimately form a sustainable feedback baseline structure. The specific implementation process includes the following steps: In the preceding steps, a time-continuous dynamic gating interval was established based on the hierarchical metabolic map. When the gating interval trigger condition is met and remains stable for more than 5 seconds, that moment is designated as the starting point for structural intervention. Subsequently, by invoking the slope of change of all metabolic parameters, the rate of change of the perturbation path direction, and the energy density distribution inside the reactor within the current moment, a perturbation response level model is established. Based on this level model, the time step of the regulation command sequence is set to 0.1 seconds, and the total control cycle duration does not exceed 15 seconds.

[0036] The command sequence consists of three types of regulating components: jet guide vanes that control the bubble path, variable aperture aeration rings that adjust bubble generation characteristics, and circulation guide devices that push the bottom fluid upwards. The action commands corresponding to these three types of components are sorted according to response priority to form a complete structural response chain. In this process, the start time, execution duration, action amplitude, termination criteria, and expected feedback indicators of each command are clearly defined, achieving temporal decoupling and task distribution of structural adjustment actions.

[0037] In the structural adjustment sequence, the earliest responders are the jet guide vanes deployed at 12 equidistant points in the upper part of the reactor. These vanes adjust their opening angle via embedded stepper motors, controlling the direction to ±60 degrees in the vertical plane. Based on the path offset angle set in the command, each vane completes its opening angle adjustment within 0.5 seconds, with the target angle set to maintain the angle between 30 and 45 degrees with the main bubble rising path, thereby generating a stable deflection disturbance channel.

[0038] After deflection occurs, some of the bubbles that were originally rising vertically will disperse and diffuse towards the inner wall of the reactor along the new path. The density decrease rate of the diffusion path is controlled within the range of 25% to 35%, while suppressing the accumulation in the middle section of the gas path and reducing the rate at which liquid is carried upward. After the guide vanes have been in operation for 6 seconds, fine adjustments are made within ±5 degrees based on the feedback of changes in gas concentration to ensure that the direction of disturbance remains deviated from the fingerprint structure of the metabolic misalignment region, thereby weakening the conditions for maintaining misalignment and breaking the gas-liquid inversion trend.

[0039] The jet adjustment action is initiated 0.5 seconds after the guide vanes are activated. The bottom aeration ring consists of thirty-six independently controlled electronic nozzles, each equipped with an adjustable nozzle head, allowing the orifice diameter to be adjusted from 0.3 mm to 1.2 mm. During the current cycle, based on the disturbance level model, the total gas flow rate is set to decrease by 20%, the nozzle diameter is uniformly adjusted to 0.5 mm, and the jet injection mode is changed to intermittent injection. Each jet cycle lasts 2 seconds, with a jet duration of 0.8 seconds and a stop time of 1.2 seconds.

[0040] This rhythmic control method constructs a discontinuous bubble chain population, resulting in a discrete distribution of bubbles in the water rather than a continuous gas column structure. This reduces the liquid-carrying capacity during ascent, avoiding the initial mechanism of gas-liquid inversion caused by excessive liquid dragging upwards. During the operation of the gas injection structure, the changes in bubble path length and path density are continuously monitored. If the deviation exceeds ±15%, the jet rhythm frequency or orifice size is automatically adjusted to achieve dynamic matching control, further enhancing the accuracy of response to local disturbances.

[0041] After the guide vanes and aeration rings have been operating in tandem for 6 seconds, the second stage of structural reconfiguration begins, triggering the bottom circulation guide device. This device consists of two sets of symmetrically arranged low-speed propellers, each 80 mm long and 100 mm in diameter. After startup, the rotational speed is controlled at 90 revolutions per minute, creating a ring-shaped tumbling flow field with a diameter of approximately 300 mm in the horizontal direction.

[0042] After the propeller starts, it pushes the liquid rich in residual substrate in the bottom region from both sides to the central axis, guiding it to tumble upwards along the central axis. This drives the underlying microbial community to move towards the metabolically active area, while simultaneously dispersing the short-chain fatty acids and dissolved gases accumulated in the middle radially along the outer ring. The morphology of this tumbling flow field is derived from the previous data on bubble path density and liquid viscosity difference, forming a closed loop. After this loop has been running stably for 8 seconds, based on the shape of the energy density equipotential surface and the uniformity of the disturbance vector distribution, the spatial field formed in this stage is set as the "steady-state feedback datum." All subsequent adjustment commands are referenced to this datum, ensuring the spatial consistency of the control strategy and the continuity of energy transfer.

[0043] The phase traction control module generates a phase conjugate respiratory traction command set under the support of the steady-state feedback base plane. Based on the residual density feedback signal and the gating signal, it performs multi-layer dynamic regulation of time-frequency inverse diffusion suppression, amplitude limiting write-back, self-suppressed interference field and anti-phase perturbation. Through phase traction and rhythm modulation of energy channels, it realizes gas-liquid partition resonance dissipation, metabolic misalignment self-recovery and steady-state closed-loop control of the system.

[0044] To achieve stable operation of the gas-liquid disturbance energy channel throughout its entire lifecycle and avoid path deviations and reactor nonlinear fluctuations caused by long-term accumulation of metabolic misalignment, it is necessary to rely on the constructed steady-state feedback baseline to further generate a dynamic response phase traction command set, and link the residual density and gating feedback signal to gradually complete the phase synchronization modulation and periodic structural correction of the energy path, ultimately constructing a closed-loop control channel. The specific implementation process is divided into the following steps: In the preceding steps, the closed-loop energy path has been completed through the jet guiding device, variable-orifice aeration structure, and lower circulation drive device, establishing a steady-state feedback baseline. Using this baseline as the original reference, the spatial coordinates, frequency distribution, and energy density change rate of the disturbance centers in different levels of the reactor are sequentially acquired and integrated into a three-dimensional path disturbance information block. For each disturbance center, the recurring jump nodes between the maximum and minimum frequency values ​​are extracted on the time axis to construct a single-path disturbance phase structure diagram. Further, using this diagram as the core, the synchronization period ratio between the upward path, the return path, and the stirring nodes is set to construct a phase coupling matrix. This matrix is ​​used to generate a phase traction command set with the three-dimensional disturbance center as the anchor point and frequency and path change values ​​as the main tuning parameters. Each command set includes five data items: spatial force application position, duration of action, disturbance frequency response range, and phase synchronization target angle, guiding the control end to form a consistent periodic disturbance rhythm within the corresponding path segment. The instruction set is iterated every 5 seconds by the internal self-operating unit to ensure that the traction action and path change are synchronized in real time. Finally, a dynamic and controllable phase alignment mechanism is established in the structural path space to realize the phase evolution of energy-guided flow at the gas-liquid interface.

[0045] After the phase traction command set starts working, the actual path amplitude, disturbance frequency, and main channel leakage energy value within the energy channel are collected every 2 seconds through the high-frequency monitoring node. The difference is calculated in real time with the phase-synchronized target trajectory to generate a "disturbance residual density distribution map". When the distribution map shows that the amplitude exceeds the set fluctuation tolerance (based on the upper limit of the main path fluctuation rate of 10%) or the frequency jumps significantly (more than twice the historical maximum frequency jump value) in any path segment, it is considered that a "reverse diffusion" anomaly has occurred in that area, which may cause local path breakage or abrupt change in metabolic parameters.

[0046] In response to this anomaly, immediately perform the following three control measures: 1) Reverse diffusion blocking intervention: The disturbance signal in the path segment is instantaneously reversed, and a set of micro-perturbation beams with the same frequency as the original signal but opposite direction are introduced. The beams are introduced from the opposite direction by the reaction component, so that the beams and the original signal cancel each other out at the intersection point, thereby achieving active interception of the reverse diffusion path. 2) Amplitude correction write-back: At the node where the anomaly occurs, extract all frequency-amplitude comparison records of the path segment within the previous 30 seconds, establish a minimum variation model, and force the current signal curve to fit the model to generate an alternative disturbance trajectory. The disturbance signal that maintains the same amplitude but does not have a sudden change is output through the servo structure, thereby keeping the energy path stable. 3) Frequency superposition interference suppression: When the main frequency of the path rises abnormally, auxiliary micro-perturbation beams with an amplitude of 70% and a frequency difference of ±5% are introduced on both sides to form a "modulation-sub-modulation" structure. The main frequency change is dispersed into multiple low-frequency interference bands through the frequency interference mechanism, thereby automatically flattening the energy amplitude and achieving the purpose of stable path control.

[0047] After the path control command set has been running for more than one cycle (set to 30 seconds), a complete path scan will be performed on all main and branch channels of the gas-liquid disturbance path. The path scan, based on the generated energy density map and frequency distribution map, evaluates the energy superposition rate, disturbance phase coordination, and response stability of each segment, and automatically extracts the residual model of typical path parameters through a gradient evaluation algorithm. This model will be used to determine whether a rhythm update is needed in the next cycle. If the three key parameters (amplitude residual, path misalignment rate, and frequency jump rate) remain within the set thresholds (e.g., amplitude residual less than 8%, misalignment rate less than 5%, and frequency jump rate less than 3%), the original control cycle will be maintained; otherwise, rhythm modulation optimization will be initiated.

[0048] The optimization methods include adjusting the traction command cycle (extending it from 5 seconds to 7 seconds), updating the reverse diffusion reverse wave structure parameters (fine-tuning the phase angle from 180 degrees to the 170-190 degree range), and replanning the interference frequency structure to make the next round of interference suppression more consistent with the dynamic evolution trend of the residual model. This adjustment process is automatic and iterative, without relying on manual input or fixed parameter settings, and has self-adaptability and high responsiveness. Ultimately, it forms a closed-loop control network with energy path phase alignment as the foundation, residual regulation as the central mechanism, and rhythm optimization as the self-adjustment method.

[0049] This invention achieves precise positioning of the anti-phase trigger point and traceability of flow field evolution through high-precision acquisition of nanosecond-level time baselines and liquid phase shear trajectories. By establishing a causal chain between bubble rising paths and energy density, it can identify nucleation regions of gas-liquid inversion in advance, preventing their expansion into systemic turbulence. Through three-dimensional vector flow field and density-viscosity coupling analysis, it constructs multi-channel disturbance correction anchor points, providing high-resolution references for monitoring spatial metabolic structures. Furthermore, through dynamic acquisition of differences in metabolic parameters between upper and lower layers, it achieves automatic identification and gating response of the layered metabolic state inside the reactor. At the structural level, it introduces jet guide vanes, variable-orifice aeration rings, and bottom guides in synergistic linkage to establish a stable feedback datum based on the restoration of energy path continuity. Finally, through a phase conjugate traction control mechanism, it achieves rhythmic suppression, phase locking, and path correction of disturbance signals, constructing a self-regulating mechanism that can adapt to dynamic environmental changes. The overall technical solution possesses complete sensing, analysis, response, and control capabilities, improving the accumulation efficiency of short-chain fatty acids, the stability of reactor operation, and the adaptability to high-load disturbance environments during co-fermentation.

[0050] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A system for co-fermenting sewage sludge with kitchen waste to produce short-chain volatile fatty acids, characterized in that, It includes a time phase construction module, a causal playback recognition module, a flow field mapping correction module, a metabolic gating feedback module, a structural reconstruction steady-state module, and a phase traction control module: The time phase construction module acquires a nanosecond-level unified time baseline, gas rise rate sequence, and liquid shear gradient trajectory. It reconstructs the gas-liquid phase difference reference surface based on the real-time flow field response and establishes a reference surface to calibrate the anti-phase trigger window. The causal playback identification module establishes a causal playback chain based on the gas-liquid phase difference reference surface, performs time reversal on the aeration rhythm, extracts the bubble rising path and flow velocity deviation, calculates the bubble rising energy density, determines the initial nucleation region of gas-liquid inversion, and generates a dynamic risk index. The flow field mapping correction module constructs a three-dimensional flow field mapping under dynamic risk index constraints, couples the gas-liquid density difference and liquid viscosity gradient to form a drift vector field, analyzes the distribution of the floating loop and the return channel, extracts the flow disturbance center and generates flow correction anchor points. The metabolic gating feedback module collects the upper layer acidity, lower layer substrate concentration and dissolved gas content of the reactor based on the flow correction anchor point, generates a stratified metabolic fingerprint, forms a metabolic stratification map based on the differences in metabolic characteristics, and establishes a threshold triggering gating interval. The structural reconfiguration steady-state module generates an adaptive structural reconfiguration sequence based on the threshold-triggered gating interval, driving the jet guide vane, variable orifice aeration ring and bottom circulation guide to respond collaboratively under time constraints, restoring the continuity of the gas-liquid reflux link and forming a steady-state feedback base surface; The phase traction control module generates a phase conjugate breathing traction command set under the support of the steady-state feedback base plane, and performs dynamic regulation of time-frequency inverse diffusion suppression, amplitude limiting write-back, self-suppressed interference field and anti-phase perturbation based on the residual density feedback signal and gating signal.

2. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 1, characterized in that, The process of reconstructing the gas-liquid phase difference reference surface is as follows: A stable frequency signal is output by an atomic clock, which is then distributed via phase-locked loop to a gas phase imaging device, a liquid phase velocity measuring device, and a time synchronization line to form a unified time baseline. The signal is then distributed to multiple fixed locations in the reactor via optical fiber to establish a consistent triggering point. Three-dimensional spatial positioning is achieved using laser illumination and a high-precision scale plate. The bubble rising velocity sequence and liquid shear gradient trajectory are obtained by high-speed imaging and particle image velocimetry, respectively. The collected data are aligned according to a unified time baseline. A phase difference analysis relationship is established between the gas rising velocity and the liquid shear velocity within the same voxel. Voxel regions with zero phase difference and consistent change direction are selected to form a gas-liquid phase difference reference surface and calibrate the anti-phase trigger window. Extract the time points, position coordinates, normal vectors, and curvature parameters that repeatedly appear within the reference plane, and establish a standard time anchor point table as a unified reference for flow field playback and structural control.

3. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 2, characterized in that, The steps for generating a dynamic risk index are as follows: A time-reversal sequence is established based on the gas-liquid phase difference reference surface and time anchor point. The bubble trajectory and liquid shear velocity of multiple cycles are aligned to form a voxel-level flow field data chain. Within the key analysis section, the velocity vector of each bubble in the voxel is calculated and compared with the liquid velocity field to obtain the relative velocity. The energy density of the bubble path is calculated based on the bubble diameter, surface tension, liquid density and liquid viscosity. Regions with energy density more than twice the average and bubble trajectory overlap exceeding a preset threshold are selected as high-disturbance candidate regions, and dynamic risk scores are generated by combining the relative position of the reference surface and the amplitude of shear velocity fluctuations. High-disturbance voxel regions with stable risk scores over multiple periods are numbered and located, and the initial nucleation region coordinates, risk index and distance to the reference surface are output to generate a dynamic risk index list, which is used as a constraint condition for flow field reconstruction.

4. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 3, characterized in that, In the calculation of bubble path energy density, the velocity difference of each bubble in each voxel is combined with the corresponding bubble diameter, liquid density, liquid viscosity and surface tension parameters, and a three-point weighted sliding window is used to smooth the velocity curve.

5. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 3, characterized in that, The process of generating flow correction anchor points is as follows: High-risk voxel regions were screened based on the dynamic risk index, a local drift vector field was constructed, and data on gas-liquid velocity difference, density difference, and viscosity gradient were recorded. By tracing the synthetic drift vector path, constructing a three-dimensional uplift and backflow channel, and identifying abnormal areas of path density and velocity change as flow imbalance zones; In the flow imbalance region, cluster strong perturbation voxels, and combine gas-liquid phase difference oscillation, shear rate abrupt change and path overlap density to screen perturbation centers; Extract the geometric center of the disturbance center and its shell stabilization structure, establish flow correction anchor points, and record their voxel numbers, triaxial velocity statistics, and shear distribution; Multiple monitoring points were set around the flow correction anchor point to collect data on acidity, dissolved gas concentration, and liquid phase temperature, constructing a metabolic profile and forming a metabolic spatial reference atlas based on the flow correction anchor point.

6. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 5, characterized in that, The steps for establishing a threshold-triggered gating interval are as follows: A five-point observation array with equal spacing was set up vertically around the identified flow correction anchor point. At each point, data on acidity, substrate concentration and dissolved gas content were collected, and metabolic data points in a three-dimensional coordinate system were formed by synchronously unifying the time baseline. The data from each observation point are used to construct acidity surface, substrate surface and gas surface, and superimposed to generate a time frame raster of metabolic map to identify metabolic misalignment regions with obvious spatial jumps. Time alignment analysis was performed on acidity abrupt changes, substrate fluctuations and gas concentration anomalies in metabolic misalignment regions, and common time periods of change were extracted to construct dynamic gating threshold intervals. When any observation array data continuously enters the gating threshold range, the structural adjustment action is triggered and the metabolic response results are recorded, based on the gas-liquid path concentration, energy density changes and abrupt changes in flow field direction, and the control parameters for the next cycle are updated.

7. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 6, characterized in that, The formation process of the steady-state feedback datum is as follows: Based on the dynamic gating interval triggering information, the starting point of structural intervention is set and a disturbance level model is constructed to generate a sequence of structural adjustment instructions. The jet guide vanes deployed in the upper middle part are driven in sequence to adjust the opening angle to form a deflection disturbance channel, disperse the bubble path and reduce the path density concentration, thus weakening the upward offset trend. After the guide vane is activated, the diameter of the aeration ring nozzle and the gas injection cycle are controlled according to the set rhythm to construct a discontinuous bubble chain group to reduce the liquid carrying effect. After the guide vanes and aeration rings complete coordinated adjustment, the bottom circulation guide device is activated to construct a tumbling flow field that combines upward and outward expansion, restore the continuity of the gas-liquid return path, and establish a steady-state feedback base.

8. The system for co-fermentation of sludge and kitchen waste to produce short-chain volatile fatty acids according to claim 7, characterized in that, The phase traction control module generates a phase conjugate traction command set under the support of the steady-state feedback base plane. Based on the residual density feedback and the gating signal, it performs the following steps: inverse diffusion suppression, amplitude limiting write-back, interference suppression, and anti-phase perturbation control. Based on the steady-state feedback datum, the spatial coordinates of the disturbance center, the rate of change of energy density and the frequency jump node are collected. The path disturbance information block is constructed and the phase traction command set is generated. The action position, action duration and synchronization target angle are set. Based on the phase traction command set, the perturbation rhythm of the path segment is controlled and the path amplitude, perturbation frequency and leakage energy in the energy channel are monitored simultaneously. The perturbation residual density distribution map is constructed and the reverse diffusion blocking intervention, amplitude correction write-back and frequency interference suppression operations are performed. After the instruction set runs, it periodically scans the energy path and generates residual models of amplitude residual, path misalignment rate and frequency jump rate. If the set threshold is exceeded, it automatically performs rhythm optimization and updates the interference structure parameters to form a closed-loop control network.

Images